Lab: The Human-Robot
Interaction Laboratory
The Human-Robot Interaction Lab (HRILab) at Tufts University aims to
enable natural interactions with humans and robots. That includes
being able to instruct robots in natural language and having
task-based dialogues with robots about their performance and their
decisions. Robots, in turns, should be able to learn on the fly from
humans if they are missing knowledge and should be able to explain
and justify their behavior to humans, with recourse to norms and
ethical principles. The HRILab is currently pursuing research in
fast interactive task learning, robot performance self-assessment,
and mechanisms for robot resilience, all in the context of
human-machine teaming and generally open-world HRI.
Research:Transfer Learning of Sensorimotor Knowledge for Embodied Agents
Prior work has shown that learning models of objects can allow robots to perceive object properties
that may be undetectable using visual input alone. However, learning
such models is costly as it requires extensive physical exploration
of the robot's environment and furthermore, such models are specific
to each individual robot that learns them and cannot directly be
used by other, morphologically different robots that use different
actions and sensors. To address these limitations, our current work
focuses on enabling robots to share sensorimotor knowledge with
other robots as to speed up learning. We have proposed two main
approaches: 1) An encoder-decoder approach which attempts to map
sensory data from one robot to another, and 2) a kernel manifold
learning framework which takes sensorimotor observations from
multiple robots and embeds them in a shared space, to be used by all
robots for perception and manipulation tasks.
Lab: The Assistive
Agent Behavior and Learning Lab
The goal of the AABL lab is to make robots more helpful. Using
insights from focus groups, user studies, and participatory
research, we design new algorithms for human-robot interaction (HRI)
and evaluate them in the real-life situations where robots could be
most useful.
We work at the intersection of assistive technology and social
robotics (including socially assistive robotics): we are especially
interested in robots interacting with users under-served by typical
HRI research, such as children, older adults, and people with
disabilities.
Ultimately, our work gives robots the intelligence to understand
what people want, make good decisions about how to help, and to help
appropriately in real-world interactions.
Research: Data and network science, systems and control, and optimization
theory and algorithms with applications in autonomous multi-agent
systems, Internet of Mobile Things (IoMTs), fleets of driverless
vehicles, and smart-and-connected cities
Research: Center for Engineering Education and Outreach
(CEEO)
A number of faculty at the Center for Engineering Education and
Outreach study the interaction between robots and
children. We look at what sort of learning results from different
ways of playing with robots, how different prompts can elicit
different different skills in the students and how designing,
building, and coding robots can provide motivation for learning more
general engineering skills. Our current research areas include
identifying the different learning catalyzed by following
instructions as opposed to providing open-ended problems in
robotics, how we can demystify AI for
K-12 students, alternate ways of teaching computational thinking
skills, multiple methods of supporting online robotic collaboration
and learning, and developing very low-cost robotic platforms for
school partners in other parts of the world. Much of our work takes
a more playful approach to learning, promoting failure and
iteration, finding personally relevant projects, and valuing and
responding to the ideas and thinking of the individual student. Our
work is funded by the NSF, LEGO Education, the LEGO Foundation, PTC,
and National Instruments.
Research: Modeling human perception for search and rescue
operations. Fusing model and sensor data to track group movements during
emergency response.
Sample paper: "Who goes there? Using an agent-based
simulation for tracking population movement"
